A connection between the number of subgrous and the order of a finite grou arxiv: v1 [math.gr] 18 Jan 019 Mihai-Silviu Lazorec January 18, 019 Abstract For a finite grou G, we associate the quantity

•2 views•18 pages

Previous

Next

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.

Documenttranscript

A connection between the number of subgrous and the order of a finite grou arxiv: v1 [math.gr] 18 Jan 019 Mihai-Silviu Lazorec January 18, 019 Abstract For a finite grou G, we associate the quantity β(g) = L(G), where L(G) is the subgrou lattice of G. Different roerties and roblems related to this ratio are studied throughout the aer. We determine the second minimum value of β on the class of -grous of order n, where n 3 is an integer. We show that the set containing the quantities β(g), where G is a finite (abelian) grou, is dense in [0, ). Finally, we consider β to be a function on L(G) and we mark some of its roerties, the main result being the classification of finite abelian -grous G satisfying β(h) 1, H L(G). MSC (010): Primary 0D30; Secondary 0D15, 0D60, 0K01. Key words: subgrou lattice, number of subgrous, finite (abelian) -grous. 1 Introduction One of the main characteristics of a finite grou G is its subgrou lattice. We denote it by L(G). The connections between G and L(G) constitute a fruitful research toic. In this regard, some interesting roblems, that were studied in the last decades, are mentioned in the Preface of the monograh [1], which is one of the well-known references on the subject. Also, the determination of the quantity L(G), where G belongs to a remarkable class of finite grous, is a roblem that is still frequently studied. As we will remark, a lot of results on this matter were obtained esecially for finite -grous. In this aer, we also focus on the quantity L(G), but we relate it to. Therefore, for a finite grou G, we study the quantity β(g) = L(G). Our aer is organized as follows. As a starting oint, in Section, we indicate some of the most relevant results concerning the counting of subgrous of a finite (abelian) -grou and we oint out some basic roerties of β. In Section 3, we determine the second minimum value of β on the class of finite -grous of order n, where n 3 is a ositive integer. As a consequence, we classify all finite -grous G or order n, with n 3, that satisfy β(g) q,n, where q,n is a quantity that deends on and n. In Section 4, we rove that the sets {β(g) G A} and {β(g) G F} are dense in [0, ), where A is the class of finite abelian grous, while F is the class of all finite grous. In Section 5, instead of studying β on a class of finite grous, we choose 1 to view it as a function on the subgrou lattice of a finite grou G. As an alication, we classify all finite abelian -grous G satisfying β(h) 1, for any subgrou H of G. Finally, some further research directions are indicated in Section 6. Mostof ournotation is standardand will usuallynot be reeated here. We only mention that we denote by L 1 (G) the oset of cyclic subgrous of a finite grou. Also, we recall that the generalized quaternion grou Q n, where n 3 is an integer, and the modular -grous M( n ), where is a rime and n is an integer such that n 3 if 3, or n 4 if =, have the following structures: Q n = x,y x n 1 = y 4 = 1,yxy 1 = x n 1 1 and M( n ) = x,y x n 1 = y = 1,x y = x 1+n. Elementary notions and results on grous can be found in []. For subgrou lattice concets we refer the reader to [1]. Preliminary results.1 Counting subgrous of finite (abelian) -grous Let G be a -grouoforder n, where n is aositive integer. Foreachintegerk such that 0 k n, we denote the number of subgrous of order k of G by s k (G). Two remarkable results involving these quantities are the following ones: Theorem.1.1. (seesection4of[1])let G be a -grou of order n. Then s k (G) 1 (mod ), k {1,,...,n 1}. Theorem.1.. (see Theorem 1 of [15]) Let G be a -grou of order n such that G = Z n and is an odd rime number. Then s k (G) 1+ (mod ), k {1,,...,n 1}. Since L(G) = + n 1 k=1 i) L(G) n+1 (mod ); s k (G), by the above results it follows that: ii) L(G) (+1)(n+1)+ (mod ); note that this congruence holds only for finite -grous G satisfying the hyotheses of Theorem.1.. There is more to say if we work with finite abelian -grous. First of all, it is well known that an abelian -grou G of order n is isomorhic to a direct roduct of cyclic -grous, i.e. G = Z d 1 Z d... Z d k, where d 1,d,...,d k areositive integerssuch that 1 d 1 d... d k and d 1 +d +...+d k = n. In other words d = (d k,d k 1,...,d 1,0,...) is a artition of n with d i 1, i {1,,...,k}, and G is a finite abelian -grou of tye d. An interesting and difficult roblem is to count the subgrous of G and an answer is given by several aers like [4, 5, 8, 11, 9]. Here, we will only recall Lemma of [8]. Let l = (l k,l k 1,...,l 1,0,...) be a artition of a non-negative integer m n, such that 0 l i d i, i {1,,...,k}. Then H l = Z l 1 Z l... Z l k is a subgrou of G of tye l. The number of subgrous of G that are isomorhic to H l is given by ] d k i=1 l k i (d k i+1 l k i+1 ) [d k i+1 l k i l k i+1 l k i, (1) where (l k,l k 1,...,l 1,0,...) and (d k,d k 1,...,d 1,0...) are the conjugate artitions (with resect to Ferrers diagrams) of l and d, resectively, while [ n k is the Gaussian binomial coefficient given ] by n ( [ nk ] i 1) = i=1. k ( i 1) n k ( i 1) i=1 It is clear that it is difficult to work with formula (1), esecially if one is interested in finding the total number of subgrous of G. Still, using different counting arguments, some exlicit formulas that allow us to obtain the quantity L(G) were given for abelian -grous of rank and 3. For more details, the reader may consult [9, 14, 18, 4, 6]. Here, we only recall that the number of subgrous of G = Z d 1 Z d, where 1 d 1 d, may be comuted using the exlicit formula L(G) = i=1 1 ( 1) [(d d 1 +1) d1+ (d d 1 1) d1+1 (d 1 +d +3)+(d 1 +d +1)], () while the number of subgrous of G = Z d 1 Z d Z d 3, where 1 d 1 d d 3, is given by where L(G) = A ( 1) ( 1), (3) A =(d 1 +1)(d 3 d +1) d1+d+5 +(d 1 +1) d1+d+4 (d 1 +1)(d 3 d ) d1+d+3 (d 1 +1) d1+d+ +(d 1 +1)(d 3 d 1) d1+d+1 (d 3 +d d 1 +3) d1+4 d1+3 + (d 3 +d d 1 1) d1+ +(d 1 +d +d 3 +5) + (d 1 +d +d 3 +1). A relevant consequence of (1) is that for all k {1,,...,n 1}, s k (G) may be viewed as a olynomial in with non-negative integer coefficients. Moreover, in his unublished work, P. Hall roved that the number of subgrous of G that are isomorhic to H l is equal to the number of subgrous of G that are isomorhic to G H l. This result is also resented on. 188 of [16] and, as a consequence, for abelian -grous G or order n we have { [ ]} n s k (G) = s n k (G), k 1,,...,. (4) Finally, we recall the main theorem of [7] which is also related to the quantities s k (G) of an abelian -grou G of order n. Theorem.1.3. Let G be an abelian -grou of order n. Then the sequence (s k (G)) k=0,[ n ] is unimodal. In other words, the last result states that the finite sequence formed of the olynomials s k (G), where k {0,1,,..., [ n ] }, has the following roerty: k { 1,,..., [ ]} n,s k (G) s k 1 (G) is a olynomial in with non-negative integer coefficients. 3 For a finite abelian -grou G or order n such that G = Z n, there are two ways to exress the quantity L(G) based on the arity of n. More exactly, if is any rime number, according to Theorem.1.1, for each k {1,,...,n 1}, there is a non-negative integer M k such that s k (G) = M k + 1. Moreover, since G is a finite non-cyclic abelian -grou, we can take M k 1, k {1,,...,n 1}, as a consequence of Proosition 1.3 of [3]. Then, using (4), we have M 1 = M n 1,M = M n,...,m [ n ] = M n [ n ]. Therefore, the number of subgrous of G is given by L(G) = { (M1 +M +...+M n 1)+(n+1), if n 1 (mod ). (5) [(M 1 +M +...+M n 1 )+M n]+(n+1), if n 0 (mod ) Similarly, if G is a -grou of order n such that G = Z n and is odd, by Theorem.1., it follows that foreach k {1,,...,n 1}, there is a non-negativeintegern k such that s k (G) = N k ++1. Following the same reasoning as the one that was done to obtain (5), we have { (N1 +N +...+Nn 1) +(n 1)(+1)+, if n 1 (mod ) L(G) = [(N 1 +N +...+Nn 1 )+Nn ] +(n 1)(+1)+, if n 0 (mod ). (6) In the end, we remark that 1 M 1 M... M [ n ] and 0 N 1 N... N [ n ], as Theorem.1.3 indicates.. Basic roerties of β Any finite grou G has at least subgrous, and, by Corollary 1.6 of [6], we have L(G) (1 4 +o(1))log, this uer bound for the number of subgrous of G being the best ossible one, in the sense that it is close to L(Z n ). It follows that β(g) ( 1 4 +o(1))log Once that increases, the lower bound goes to 0, while the uer one aroaches infinity. Hence, it is clear that β(g) (0, ). If two finite grous G 1 and G are isomorhic, then β(g 1 ) = β(g ). The converse is false since β(z Z 8 ) = β(m 16 ) = β(q 16 ) = 11 16, and any two of the above three grous are not isomorhic. An imortant roerty for our study is the multilicativity of β. This roerty lays a significant role when it comes to rove some density results related to β in Section 4. Proosition..1. Let (G i ) i=1,k be a family of finite grous having corime orders. Then k β( G i ) = i=1 4 k β(g i ). i=1. A direct consequence of this multilicativity is exressing the quantity β associated to any finite nilotent grou G since G can be written as the direct roduct of its Sylow subgrous. So, if (P i ) i=1,k are the Sylow subgrous of G, we have β(g) = k β(p i ). i=1 One may rove that any finite grou G having at most 5 subgrous is abelian. There are different ways to check this, but we give a roof which is connected with some of the results that were recalled in the revious Subsection. Also, we write the above roerty in terms of β. Proosition... Let G be a finite grou. If β(g) 5, then G is abelian. Proof. Suose that G is non-abelian. If G is a -grou of order n, since G would have at least one -subgrou of order k for each k n and G is non-abelian, it follows that n {3,4}. Then L(G) 3+s 1 (G). By Theorem.1.1, we have s 1 (G) +1 as a consequence of the fact that G is non-cyclic. Then β(g) +4 6, a contradiction. If G is not a -grou, there are exactly distinct rime numbers, say and q, that are divisors of. Indeed, if at least 4 such divisors exist, then β(g) 5 as a consequence of Cauchy s theorem, a contradiction. If there are 3 distinct rime divisors of, then G has 3 Sylow subgrous of different orders and all of them must be normal. Otherwise Sylow s 3rd theorem would imly that β(g) 5, a contradiction. But, since all normal subgrous are ermutable, again we would arrive at the same contradiction. Therefore, G has Sylow subgrous P and Q of orders x and q y, resectively, where x and y are ositive integers. Then β(g) x+y+, since P and Q have at least x and y non-trivial subgrous, resectively. If x 3 or y 3, we contradict our hyothesis. The cases (x,y) {(1,),(,1)} are excluded using the same reasoning involving Sylow s 3rd theorem and the fact that normality imlies ermutability. Then x = y = 1, so G is a grou of order q. Since G is non-abelian, it follows that P or Q is not a normal subgrou of G. Then, once again as a consequence of Sylow s 3rd theorem, a contradiction. Hence, our assumtion that β(g) 5 G is non-abelian is false and the roof is finished. 3 Bounds and minima roblems related to -grous The aim of this Section is to add some roertiesofβ by restrictingourstudy to finite -grous. We start by indicating some bounds for this quantity and we continue by solving some minima roblems, the main result ointing out the finite -grous for which β attains its second minimum value. We denote by P the class of -grous of order n, where n 3. Let G P. Then β(g) = n s k (G) n. k=0 Since G has at least one subgrou of order k, k {1,,...,n 1}, the minimum value of β, on P, is attained if and only if s k (G) = 1, k {1,,...,n 1}, i.e., if and only if G = Z n. Using 5 Theorem 5.17 of [3], we deduce that β attains its maximum value on P if and only if G = Z n. Hence, for all G P, we have β(z n) β(g) β(z n ). A connection between G P and some of its quotients may be written as a consequence of Theorem 1.3 of [0]. More exactly, let G P and H be a normal subgrou of G such that H =. Then, ( ) G β(g) β H Z. In the same aer (see Theorem 1.4), the author roves that if G is a non-elementary abelian -grou of order n, where is odd and n 3 is a ositive integer, then where s k (G) s k (M (1,1,1) Z n 3 ), k {1,,...,n 1}, M (1,1,1) = a,b,c a = b = c = 1,[a,b] = c,[c,a] = [c,b] = 1. He conjectures that this result also holds for = and a roof of this fact is given in [7]. This means that the second maximum value of β, on P, is attained when one works with the grou M (1,1,1) Z n 3. What about the second minimum value? In what follows, we rovide an answer to this question. In this regard, it is worth to recall Theorem. of [19]. This result states that for a finite -grou G of order n, we have s k (G) = + 1, k {1,,...,n 1} if and only if G = Z Z n 1 or G = M( n ). Firstly, we indicate an answer to the above question if we work only with abelian grous contained in P. Proosition 3.1. Let G P such that G is abelian and G = Z n. Then β(g) β(z Z n 1). The equality holds if and only if G = Z Z n 1. Proof. Let Gbeagrouasindicated byourhyothesisandsuosethatβ(g) β(z Z n 1). By (), we have β(z Z n 1) = (n 1)(+1)+ n. If n is odd, then according to (5), we deduce that β(g) β(z Z n 1) = (M 1 +M +...+M n 1) n 1, where 1 M 1 M... M n 1. But, (M 1 +M +...+M n 1) n 1 = n 1, and this leads to a contradiction. Consequently, the inequality β(g) β(z Z n 1) holds. Following the above reasoning, one can analyse the case of even integers n 3 and arrive at the same conclusion. 6 Concerning the situation where equality holds, we have β(g) = β(z Z n 1) M 1 +M +...+M n 1 = n 1, where M k 1, k {1,,...,n 1}. Then, β(g) = β(z Z n 1) M k = 1, k {1,,...,n 1} s k (G) = +1, k {1,,...,n 1} G = Z Z n 1 We note that the last equivalence is a consequence of the classification that was recalled above. In what follows, we find the second minimum of β on the entire P. It is quite interesting that, in some cases, there are at least minimum oints in P associated to this minima roblem. Theorem 3.. Let G P such that G = Z n. i) If is odd, then β(g) β(z Z n 1) = β(m( n )). β(g) β(q 8 ), if n = 3 ii) If =, then β(g) β(z Z 8 ) = β(q 16 ) = β(m(16)), if n = 4. β(g) β(z Z n 1) = β(m( n )), if n 4 The equality holds if and only if G is isomorhic to one of the indicated minimum oints corresonding to each case. Proof. Let G P with G = Z n. We must find another grou G 1 having the same roerties such that β(g) β(g 1 ). To obtain the smallest value of β(g 1 ), we must lower the quantities s k (G 1 ), k {1,,...,n 1}. If there is an integer k such that s k (G 1 ) = 1, since G 1 is noncyclic, by Proosition 1.3 of [3], the revious equality holds if and only if G 1 = Q n. Otherwise, if s k (G 1 ) 1, k {1,,...,n 1}, then Theorem.1.1 indicates that the lowest ossible value of s k (G 1 ) is +1 for all k {1,,...,n 1}. As we reviously remarked, this haens if and only if G 1 = Z Z n 1 or G 1 = M( n ). Hence, if is odd, then the minimum oint G 1 is isomorhic to Z Z n 1 or M( n ) and the second minimum value of β, on P, is β(z Z n 1) = β(m( n )) = (n 1)(+1)+ n. If =, the determination of the minimum oint G 1 is related to the value of n. We recall that this result being indicated in [3]. We have L(Q n) = n 1 +n 1, β(z Z n 1) = β(m( n )) = 3(n 1)+ n and β(q n) = n 1 +n 1 n. It is easy to check that β(q n) β(z Z n 1) = β(m( n )) for any integer n 4 and that the equality holds for n = 4. Then, in order to finish the roof, we distinguish the following 3 cases: 7 If n = 3, then the minimum oint is G 1 = Q8 and β(g) β(q 8 ) = 3 4. If n = 4, then the minimum oint G 1 is isomorhic to Z Z 8,Q 16 or M(16) and β(g) β(z Z 8 ) = β(q 16 ) = β(m(16)) = If n 4, then the minimum oint G 1 is isomorhic to Z Z n 1 or M( n ) and β(g) β(z Z n 1) = β(m( n )) = 3(n 1)+ n. Since we determined all ossible minimum oints G 1 corresonding to each case, the equality β(g) = β(g 1 ) holds if and only if G = G 1. The results that were roved in this Section may be also interreted as follows. Corrolary 3.3. Let G P such that G is abelian. Then β(g) (n 1)(+1)+ n Corrolary 3.4. Let G P. i) If is odd, then G = Z n or G = Z Z n 1. β(g) (n 1)(+1)+ n G = Z n,g = Z Z n 1 or G = M( n ). β(g) 3 4 G = Z 8 or G = Q 8, for n = 3 ii) If =, then β(g) G = Z 16,G = Z Z 8,G = Q 16 or G = M(16), for n = 4. β(g) 3(n 1)+ G = Z n n,g = Z Z n 1 or G = M( n ), for n 4 One may go further and try to find the third minimum (maximum) value of β, on the class of -grous of order n, where n 4. In this regard, we rove a result which may be considered as a starting oint for such a study. More exactly, we show that the third minimum value of β, on the class of abelian -grous of order n such that n 4 and is odd, is attained at the oint Z Z n. We recall that for a ositive integer n, there is a bijection between the set of artitions of n and the set of tyes of abelian -grous of order n. More exactly, for a artition d = (d k,d k 1,...,d 1,0,...) of n, there is an unique abelian -grou G = Z d 1 Z d... Z d k of tye d and order n. Moreover, the relation defined by (d k,d k 1,...,d 1,0,...) (e k,e k 1,...,e 1,0,...) d i = e i, i {1,,...,k} or i {1,,...,k 1} such that d i e i and d i+1 = e i+1,...,d k = e k, 8 is a total order on the set containing all artitions of n. We are ready to rove a reliminary result that will also be helful in Section 5. Lemma 3.5. β is strictly decreasing on the class of abelian -grous of rank and order n, where n 4. Proof. Let n 4 be a ositive integer. Without loss of generality, we choose the abelian -grou G 1 = Z d 1 Z d of tye d = (d,d 1,0,...) and order n, where d 1 d. Since the set of artitions of n is totally ordered, it is sufficient to take the consecutive artition of d with resect to, i.e. (d +1,d 1 1,0,...), and the corresonding grou G = Z d 1 1 Z d +1, and rove that β(g ) β(g 1 ). Using (), we have β(g ) β(g 1 ) (d d 1 +3) (d d 1 +1) (d d 1 +1) (d d 1 1) +1. Since the last inequality holds for any rime, the roof is comlete. We remark that, in general, β is not monotonic on the class of finite abelian -grous of a given order. For instance, if = and n = 9, then we have(4,4,1,0,...) (5,,,0,...) (5,3,1,0,...), but β(z Z 16 Z 16 ) = 3,β(Z 4 Z 4 Z 3 ) = 354 and β(z Z 8 Z 3 ) = 58. We mention that these numbers may be obtained using (3) or GAP [8]. Finally, before roving the last result of this Section, we mention that for any finite -grou G of order n, there is a bijection between the set containing all maximal subgrous of G and the set G formed of the maximal subgrous of Φ(G) = Z d(g). Here, we denoted the Frattini subgrou and the minimal number of generators of G by Φ(G) and d(g), resectively. Hence, s n 1 (G) = s d(g) 1 (Z d(g) ) = d(g) 1 1. Proosition 3.5. Let G be an abelian -grou of order n such that n 4 and is odd. Suose that G = Z n and G = Z Z n 1. Then β(g) β(z Z n ). Proof. Let G be a grou as indicated by our hyothesis. Suose that β(g) β(z Z n ). We choose n 4 to be an even integer and we mention that one may follow the same reasoning to comlete the roof for odd integers. By (), we have β(z Z n ) = (n 3) +(n 1)(+1)+ n. Therefore, since G is non-cyclic and is odd, we can use (6) to deduce that β(g) β(z Z n ) (N 1 +N +...+Nn 1 )+Nn n 3, where 0 N 1 N... Nn. If N 1 1, then (N 1 +N +...+Nn 1 )+Nn n 1, 9 a contradiction. Then N 1 = 0, so, by (4), we have s 1 (G) = s n 1 (G) = +1. But, as we mentioned above, we also have s n 1 (G) = d(g) 1 1. It follows that d(g) =. Consequently, G = Z d 1 Z d, where 1 d 1 d. According to Lemma 3.4, since β(g) β(z Z n ), it follows that d 1 = 1 and d = n 1. Then G = Z Z n 1 and we contradict our hyothesis. Therefore β(g) β(z Z n ), as desired. 4 Some density results associated to β We denote by F the class of all finite grous. Let A be the subclass of F containing all finite abelian grous. In this Section, our main aim is to rove that the sets are dense in [0, ). {β(g) G A} and {β(g) G F} A first ste to reach our urose is to recall the Proosition marked on. 863 of [17]. This result mainly states that given a sequence (x n ) n 1 of ositive real numbers such that lim x n = 0 n and x n is divergent, then ({x i } i=1 ) = [0, )

We Need Your Support

Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...Sign Now!

We are very appreciated for your Prompt Action!

x